supporting online material for -...

Originally posted 8 September 2011; corrected 29 September

www.sciencemag.org/cgi/content/full/science.1208592/DC1

Supporting Online Material for

An Expanded Palette of Genetically Encoded Ca Indicators

Yongxin Zhao, Satoko Araki, Jiahui Wu, Takayuki Teramoto, Yu-Fen Chang, Masahiro Nakano, Ahmed S. Abdelfattah, Manabi Fujiwara, Takeshi Ishihara, Takeharu Nagai,

Robert E. Campbell*

*To whom correspondence should be addressed. E-mail: [email protected]

Published 8 September 2011 on Science Express DOI: 10.1126/science.1208592

This PDF file includes

Materials and Methods SOM Text Figs. S1 to S9 Tables S1 to S6 References

Other Supporting Online Material for this manuscript includes the following: (available at www.sciencemag.org/cgi/content/full/science.1208592/DC1)

Movie S1. Time-lapse three-color fluorescence imaging of HeLa cells transfected with plasmids encoding R-GECO1 targeted to the nucleus (top right), G-GECO1 targeted to the cytoplasm (top left), and GEM-GECO1 to the mitochondria (bottom left). Filter specifications are provided in Table S5. For GEM-GECO1, the ratio of blue to green fluorescence emission has been pseudocolored by using the color bar shown. (Bottom right) Image merges the three color channels, with the GEM-GECO1 emission ratio pseudocolored magenta rather than the rainbow representation used in the adjacent panel.

Corrected 29 September 2010: Four revisions were made by the authors: on p. 4, “YY” was deleted and replaced with “RV_GCaMP-Stop-HindIII”; on p. 8, “25°C” was changed to “20°C”; on p. 10, “GCaMP_RV_EcoR1” was changed to “GCaMP_RV_HindIII”; and in Table S6, a row was added for GCaMP_RV_HindIII.

of 37

General

Synthetic DNA oligonucleotides used for cloning and library construction were purchased from

Integrated DNA Technologies or Hokkaido System Science. The sequences of all oligonucleotides

used in this work are provided in Table S6. Pfu polymerase (Fermentas) or KOD+ (Toyobo Life

Science) were used for non-mutagenic PCR amplifications in the buffer supplied by the respective

manufacturer. Taq polymerase (New England Biolabs) in the presence of MnCl2 (0.1 mM) was used

for error-prone PCR amplifications. PCR products and products of restriction digests were routinely

purified using preparative agarose gel electrophoresis followed by DNA isolation using the GeneJET

gel extraction kit (Fermentas) or QIAEX II gel extraction kit (Qiagen). Restriction endonucleases were

purchased from Fermentas or Takara Biosciences and used according to the manufacturer’s

recommended protocol. Ligations were performed using T4 ligase in Rapid Ligation Buffer (Promega).

Small-scale isolation of plasmid DNA was performed by GeneJET miniprep kit (Fermentas) or by

alkaline lysis of bacteria (a pellet derived from 1.5-3 mL of liquid culture) followed by ethanol

precipitation of the DNA. Large-scale plasmid DNA purifications were performed by alkaline lysis of

bacteria (a pellet derived from 150-200 mL of liquid culture) followed by successive steps of

isopropanol precipitation, PEG 8000 precipitation, and 2 rounds of phenol/chloroform extraction. The

cDNA sequences for all GECOs and fusion constructs were confirmed by dye terminator cycle

sequencing using the BigDye Terminator Cycle Sequencing kit (Applied Biosystems). The source and

detailed specifications for all filters used for fluorescence screening and imaging are provided in Table

S5. To determine the Ca2+ concentration in growth media, LB liquid medium was prepared according

to manufacturer’s instruction (Broth Lennox, Fisher Scientific, catalog number BP1427-2) and purified

GEM-GECO was used to determine the concentration of free Ca2+ using a standard addition method to

account for matrix effects.

Construction of the TorA protein export plasmid for GCaMP3

of 37

The TorA protein export plasmid (pTorPE) was constructed by inserting a digested DNA fragment

encoding TorA-6×His-GCaMP3-SsrA into the NcoI/HindII sites of pBAD/His B (Invitrogen). The

insert region was assembled by a 2-part overlap extension PCR. The 5’ piece used in the overlap

extension contained the DNA encoding the TorA signal peptide and was prepared by PCR

amplification of E. coli strain DH10B (Invitrogen) genomic DNA with primers FW_TorA and

RV_TorA. Primer FW_TorA contains a NcoI site and primer RV_TorA contains a XbaI site and a

region that overlaps with primer FW_XbaI-6His1. The 3’ piece used in the overlap extension was

prepared by PCR amplification of the gene encoding 6×His-GCaMP3 (in vector backbone pEGFP-N1,

Addgene plasmid 22692) with primers FW_XbaI_6His1 and RV_SsrA-GCaMP3. Primer RV_SsrA-

GCaMP3 contains an XmaI site followed by the coding sequence for the SsrA tag and a HindIII site.

Prior to PCR amplification, a NcoI site in GCaMP3 domain was removed using the Quikchange

method with primer FW_GCaMP3_c105a. The PCR fragments were purified by agarose gel

electrophoresis. The full length TorA-6×His-GCaMP3-SsrA chimera was assembled by overlap

extension PCR using primers FW_TorA and RV_SsrA-GCaMP3 and a template composed of a

mixture of the 5’ and 3’ PCR fragments (1 μL each). The full-length product (approximately 1400 bp)

was purified by agarose gel electrophoresis and the doubly digested product was ligated into the NcoI

and HindIII sites of pBAD/His B (Invitrogen) to yield pTorPE-GCaMP3 with NcoI-TorA-XbaI-6×His-

GCaMP3-XmaI-SsrA–HindIII in the multiple cloning site.

Construction of GECO gene libraries

Error-prone PCR amplifications for construction of libraries of randomly mutated genes was performed

as described above. In the first two rounds of G-GECO library construction, primers FW_XbaI-6His2

and RV_GCaMP-XmaI (in which an XmaI site follows the last residue of the CaM portion of the

GECO construct) were used for error-prone PCR. The resulting PCR products were digested with Xba1

of 37

and Xma1 and ligated with similarly digested pTorPE-GCaMP3. Subsequent to our observation that

the SsrA tag was truncated in some of our most favorable variants, we switched to using reverse primer

RV_GCaMP-Stop-HindIII in which a stop codon and a HindIII site immediately follow the last residue of

the CaM. In this case, the PCR products were digested with Xba1 and HindIII and ligated into the

similarly digested pTorPE vector backbone. For site-directed mutagenesis or library construction by

full or partial randomization of one or more codons, either the QuikChange Multi kit (Agilent

Technologies) or overlap extension PCR was employed. Following ligation, electrocompetent E. coli

strain DH10B cells was transformed with the library of gene variants and cultured overnight at 37 � on

10-cm Petri dishes of LB-agar supplemented with 50 μg/mL ampicillin (Sigma) and 0.0002% to 0.0008%

(wt/vol) L-arabinose (Alfa Aesar). The initial B-GECO library was constructed using the QuikChange

Multi kit with G-GECO1.1 as template and primers G-B_V63VILM, G-B_T223ST_Y224H and G-

B_R377X_N380X.

The initial cpmApple gene library was assembled by a two-part overlap extension PCR. The 5’

piece used in the overlap extension was prepared by PCR amplification with a mixture of two different

forward primers (FW_XhoI-X-147mc, FW_XhoI-X-148mc) and a single reverse primer

(RV_GGTGGS-mCherry). Primers FW_XhoI-X-147mc and FW_XhoI-X-148mc contains a XhoI site

and primer RV_GGTGGS-mCherry contains a cp linker GGTGGS and an overlap region with primer

FW_GGTGGS-mCherry. The 3’ piece for use in the overlap extension was prepared by PCR

amplification with using a single forward primer (FW_GGTGGS-mCherry) and a mixture of two

different revers primers (RV_MluI-X 143mc and RV_MluI-X 144mc). Primer RV_MluI-X 143mc and

RV_MluI-X 144mc contains an MluI site. The PCR fragments were confirmed by agarose gel

electrophoresis and purified. The full-length cpmApple gene library was assembled by overlap

extension PCR using an equimolar mixture of primers FW_XhoI-X-147mc, FW_XhoI-X-148mc,

RV_MluI-X 143mc and RV_MluI-X 144mc together with a mixture of the 5’ and 3’ PCR fragments (1

of 37

μL each) as the template. The full-length product (approximately 1400 bp) was purified by agarose gel

electrophoresis and the doubly digested product was ligated into the XhoI and MluI sites of a modified

pTorPE-G-GECO1.1 in which a second MluI site had been first removed by QuikChange Multi

procedure with primer Destroy_MluI.

Screening of GECO gene libraries

The imaging system used for library screening has previously been described in detail (18). For

libraries generated by random mutagenesis, we screened 10,000–20,000 colonies (10–20 plates of

bacterial colonies) per round, typically stopping screening when a number of substantially improved

variants had been identified. For libraries generated by randomization of one or more codons, we

screened approximately 3-fold more colonies than the expected library diversity (e.g., 12,000 colonies

for a 4,000-member library). During library screening we picked colonies that exhibited the top 0.01%

to 0.1% change of fluorescence intensity (or ratio) upon application of an EGTA solution (2 mM

EGTA, 30 mM Tris-HCl, pH 7.4; applied using a spray bottle that produced a fine mist). In later

rounds of screening, we also picked the top 0.01% to 0.1% of colonies based only on their brightness

(or ratio) prior to application of EGTA. Picked clones were individually cultured in 4 mL liquid LB

medium (0.0016% L-arabinose, 200 μg/ml ampicillin) that was then shaken (250 rpm) for either 36 h at

30 °C (in earlier generations) or 22 h at 37 °C (in later generations).

Proteins extracted from the liquid cultures of the picked clones were subjected to a second stage

of screening in a Safire2 fluorescence microplate reader (Tecan). Extraction of periplasmic protein

from E. coli cultures was performed by a cold osmotic shock procedure. Briefly, bacterial cells were

harvested by centrifugation at 13,000 g for 2 min at 4 °C and gently resuspended in 500 μL of ice-cold

pH 8.0 buffer containing 30 mM Tris·Cl, 1 mM EDTA and 20% sucrose. After 5 min of gentle

agitation on ice, the bacteria were again pelleted by centrifugation (9,000 g for 5 min at 4°C),

of 37

resuspended in 400 μL of ice-cold 5 mM MgSO4, and gently agitated for 10 min on ice. Following

centrifugation to pellet the intact bacteria (9,000 g for 5 min at 4°C), the supernatant (the osmotic

shock fluid containing the periplasmic protein fraction) was collected. In cases where the periplasmic

export efficiency was particularly low, cytoplasmic protein was extracted by suspension of the osmotic

shock cell fraction in B-PER (Pierce) followed by centrifugation to pellet the cell debris.

Protein purification and in vitro spectroscopy

To purify GECOs for in vitro spectroscopic characterization, the pTorPE plasmid harboring the variant

of interest was first used to transform electrocompetent E. coli DH10B cells. Following selection on

LB/ampicillin (200 μg/ml), single colonies were picked and used to inoculate 4 mL LB medium (200

μg/ml ampicillin, 0.0016% L-arabinose). Bacterial cultures were shaken at 250 rpm and allowed to

grow for 40 h at 30 °C. Bacteria were harvested by centrifugation (10,000 g for 5 min), resuspended in

30 mM Tris-HCl buffer (pH 7.4), lysed by French press, and clarified by centrifugation at 13,000 g for

45 mins at 4°C. Proteins were purified from the cell-free extract by Ni-NTA affinity chromatography

(Agarose Bead Technologies). Purified proteins were dialyzed into either 30 mM Tris, 150 mM NaCl,

pH 7.4 or 10 mM MOPS, 100 mM KCl, pH 7.2. Absorption spectra were recorded on a DU-800 UV-

visible spectrophotometer (Beckman) and fluorescence spectra were recorded on either a Safire2

platereader (Tecan) or a QuantaMaster spectrofluorometer (Photon Technology International). For

ratiometric GECOs, the response to Ca2+ is expressed as 100 * (Rmax-Rmin)/Rmin. For GEM-GECO, R =

(I at 455 nm)/(I at 511 nm), where I = fluorescence intensity when excited at 400 nm. For GEX-GECO

R = (I with 400 nm excitation)/(I with 488 nm excitation), where I = fluorescence intensity at 510 nm.

Standards for quantum yield determination were quinine (for B-GECO, GEX-GECO, and GEM-

GECO), fluorescein (for G-GECO), and mCherry (for R-GECO). Briefly, the concentration of protein

in a buffered solution (30 mM MOPS, pH 7.2, with either 10 mM EGTA or 10 mM Ca-EGTA) was

of 37

adjusted such that the absorbance at the excitation wavelength was between 0.2 and 0.6. A series of

dilutions of each protein solution and standard, with absorbance values ranging from 0.01 to 0.05, was

prepared. The fluorescence spectra of each dilution of each standard and protein solution were recorded

and the total fluorescence intensities obtained by integration. Integrated fluorescence intensity vs.

absorbance was plotted for each protein and each standard. Quantum yields were determined from the

slopes (S) of each line using the equation: Φprotein = Φstandard × (Sprotein/Sstandard).

Extinction coefficients were determined by first measuring the absorption spectrum of each

GECO in Ca2+-free buffer (30 mM MOPs, 100 mM KCl and 10 mM EGTA at pH 7.2) and Ca2+-buffer

(30 mM MOPS, 100 mM KCl and 10 mM Ca-EGTA at pH 7.2). The concentrations of G-GECO,

GEM-GECO and GEX-GECO were determined by measuring the absorbance following alkaline

denaturation and assuming ε = 44,000 M-1cm-1 at 446 nm. For R-GECO, the protein concentration was

determined by comparing the absorption peak for denatured R-GECO to that of denatured mCherry

(determined to have ε = 38,000 M-1cm-1 at 455 nm) following alkaline denaturation of both proteins.

The concentration of B-GECO was determined by BCA assay (Pierce). Extinction coefficients of each

protein were calculated by dividing the peak absorbance maximum by the concentration of protein.

To determine the apparent pKa for each GECO, a series of phosphate-free buffers was prepared as

follows. A solution containing 30 mM trisodium citrate and 30 mM borax was adjusted to pH 11.5 and

HCl (12 M and 1M) was then added dropwise to provide solutions with pH values ranging from 11.5 to

3 in 0.5 pH unit intervals. The pH titration of Ca2+-free protein were performed by adding 1 μL of

concentrated protein in Ca2+-free buffer (30 mM MOPS, 100 mM KCl, 10 mM EGTA, at pH 7.2) into

100 μL of each of the buffers described above. The pH titration of the Ca2+-bound protein was

performed by adding 1 μL of protein in Ca2+ containing buffer (30 mM MOPS, 100 mM KCl and 10

of 37

mM CaCl2, pH 7.2) into 100 μL of the pH buffers. The fluorescence of each GECO in each buffer

condition was recorded using a Safire2 multiwell fluorescence platereader (Tecan).

Determination of Ca2+ Kd’ and Ca2+-association kinetics

Ca2+ titrations were performed by dilution (1:100) of a concentrated protein solution into a series of

buffers which were prepared by mixing Ca2+-saturated and Ca2+-free buffers (30 mM MOPS, 100 mM

KCl, 10 mM EGTA, pH 7.2, either with or without 10 mM Ca2+) to provide a series of solutions with

free Ca2+ concentration ranges from 0 nM to 3,900 nM at 20 °C (19). The fluorescence intensity of

protein in each solution was determined and plotted as a function of Ca2+ concentration. Experiments

were performed in triplicate and the averaged data from the three independent measurements was fit to

the Hill equation.

Ca2+-association kinetics were determined by stopped-flow photometry on a SX20 stopped-flow

spectrometer (Applied Photophysics). The GECO indicator (in 30 mM MOPs, 1 mM EGTA and 100

mM KCl) was rapidly mixed (1:1) with a series of Ca2+ buffers that were prepared by mixing of a

buffered solution (30 mM MOPs, 100 mM KCl) with different ratios of 10 mM EGTA and 10 mM Ca-

EGTA. The change in the fluorescence or absorbance signal during rapid mixing provided the

relaxation rate constants (kobs) for the Ca2+ association reaction at various Ca2+ concentrations (from

300 nM to 1500 nM). By fitting the observed data to the equation kobs =kon × [Ca2+]n + koff, the kinetic

constants of Ca2+ association kon and dissociation koff were determined.

Construction of mammalian expression plasmids.

To construct the plasmids for validation and testing of Ca2+ indicators in HeLa cells, the gene in the

pBAD vector was first PCR amplified using primers GCaMP_FW_BamH1 and GCaMP_RV_EcoR1.

For B-GECOs that contained the Ser5Pro substitution, primer Cpd_S5P_FW_BamH1 was used in

place of GCaMP_FW_BamH1. PCR products were subjected to phenol/chloroform extraction followed

of 37

by ethanol precipitation, then redissolved in 16 μL of deionized water. The DNA was digested with

BamH1 and EcoR1 in a total volume of 20 μL and then purified by electrophoresis on a 1% agarose gel

(Agarose S, Nippon Gene Co.) followed by gel extraction. Purified DNA fragments were ligated into

the polylinker region of a modified pcDNA3 plasmid digested with BamH1 and EcoR1 and similarly

purified. The modified pcDNA3 plasmid had been made by deleting 2224 nucleotides (including the

SV40 promoter, the SV40 origin of replication, the Neomycin ORF, and the SV40 poly A region) from

the original 5.4 kb pcDNA3. The ligation reaction was used for transformation of chemically

competent E. coli XL10-Gold (Agilent Technologies) by heat shock at 42 � for 45 s. Cells were plated

on LB/agar supplemented with ampicillin and individual colonies were picked into 3 mL of

LB/ampicillin following overnight incubation at 37 �. Liquid cultures were shaken at 155 rpm and 37 �

for 12-15 h and then a small scale isolation of plasmid DNA was performed. Each gene was fully

sequenced using T7_FW, F-EGFP-C, and BGH_RV primers. Once the sequence had been confirmed,

E. coli XL10-Gold was again transformed with each plasmid, plated, and an individual colony was

used to inoculate 150-200 mL of LB/ampicillin. Following shaking at either 37 � for 12-15 h or 23 �

for 48 h, a large-scale plasmid purification was performed on the culture. Plasmids from large-scale

purifications were stored in TE buffer and used for transfection as described below.

To construct the nucleus-targeting expression plasmids for use in multicolor imaging, a similar

procedure to that described above was used. Briefly, a two-step PCR protocol was used. In the first

PCR step, the GECO of interest was PCR amplified with GCaMP_RV_EcoR1 and either

GCaMP_2NLS_FW1 or Cpd_2NLS_FW1 (for B-GECO). The resulting PCR product is used as the

template in a second PCR amplification with GCaMP_RV_EcoR1 and Cpd_2NLS_FW2. The full-

length PCR product, which contains the gene encoding the GECO of interest, with 2 copies of the NLS

sequence (DPKKKRKV) at the N-terminal end, was ligated into the BamH1 and EcoR1 sites of the

modified pcDNA3 plasmid as described above.

of 37

To construct the mitochondria-targeting expression plasmid, pcDNA3-mitAT1.03 (14) was

digested with BamHI and EcoRI and the plasmid backbone purified by gel electrophoresis. To prepare

the insert for ligation with the cut plasmid, the GECO of interest was PCR amplified using

GCaMP_RV_HindIII and either GCaMP_FW_BamH1_mito or Cpd_S5P_FW_BamH1_mito (for B-

GECO). Digestion and purification of the insert DNA, followed by ligation with the digested plasmid

afforded an expression vector encoding the vector of interest fused to 2 copies of the mitochondria

targeting sequence of cytochrome c oxidase subunit VIII at the N-terminal end.

HeLa cell culture imaging

HeLa cells (40-60% confluent) on collagen-coated 35 mm glass bottom dishes (Mastumami) were

transfected with 1 μg of plasmid DNA and 4 μL lipofectamine 2000 (Invitrogen) according to the

manufacturers instructions. After 2 h incubation the media was exchanged to DMEM with 10% fetal

bovine serum and the cells were incubated for an additional 24 h at 37 � in a CO2 incubator. For flash-

pericam variants, cells were incubated at 28 �. Immediately prior to imaging, cells were washed twice

with Hanks balanced salt solution (HBSS) and then 1 mL of 20 mM HEPES buffered HBSS was added.

Cell imaging was performed with an inverted Eclipse Ti-E (Nikon) equipped with an Orca-R2

digital CCD camera (Hamamatsu Photonics K.K.), a Micro scanning stage MSS-BT110 (Chuo

Precision Industrial Co. Ltd.), and a Chamlide IC incubator system (Live Cell instrument). The

AquaCosmos software package (Hamamatsu Photonics K.K.) was used for automated microscope and

camera control. For determination of dynamic ranges in live cells, cells were imaged with a 40× oil

objective lens (NA 1.3), excitation light intensity was decreased to 1% with a neutral density filter

(ND100), and a 0.6× lens was placed between the microscope and the CCD camera. All imaging, with

the exception of cells expressing flash-pericam variants, was performed at 37 �. Flash-pericam

of 37

variants were imaged at room temperature. For 3-color GECO imaging, an objective lens with

excellent chromatic aberration correction through the visible range is strongly recommended.

For imaging of histamine-induced Ca2+ dynamics, cells were imaged with a 500 ms exposure

(2×2 binning) acquired every 5 s for a duration of 15 min. Approximately 30 s after the start of the

experiment, histamine (10 μL) was added to a final concentration of 5 μM. Once the experiment had

ended, cells were washed 2× with HBSS, and then incubated for 10 min in 1 mL HHBSS to allow

histamine-induced oscillations to subside. Cells are then imaged as described above, with exposures

every 10 s for a duration of 5 min. Approximately 30 s after imaging is started, 1 mL of 2 mM CaCl2,

10 μM ionomycin in Ca2+- and Mg2+-free HHBSS (HHBSS(-)) is added to the dish via peristaltic pump.

At the end of the experiment, cells are washed 3× with HHBSS(-). Following addition of 1 mL of 1

mM EGTA, 5 μM ionomycin in HHBSS(-), cells were imaged again with exposures every 5 s for a

total of 5 min.

For co-imaging of GECOs and fura-2, HeLa cells were transfected and cultured 24 h at 37 � as

described above. Cells were then incubated for 15 min at room temperature with 1 μM fura-2 AM and

0.5× PowerLoad concentrate (Invitrogen) in HHBSS. Cells were washed twice with HHBSS and then

imaged in HHBSS with microscope settings identical to those described above. Filters are provided in

Table S5.

In Fig. 3GH, GEM-GECO1 and GEX-GECO1 ratios are expressed on the right hand axis as pCa

= –log10[Ca2+]. Ratios were converted to Ca2+ concentrations using the equation [Ca2+] = (Kd’n×(R-

Rmin)/(Rmax-R))1/n; where Kd’ is the apparent dissociation constant (Table S2); n is the Hill coefficient

(Table S2); R is the experimental ratio; Rmin is the experimental ratio obtained during ionomycin/EGTA

treatment; and Rmax is the ratio during ionomycin/Ca2+ treatment (20).

Preparation and imaging of rat hippocampal neuron culture

All experimental procedures were performed in accordance with the guidelines of the Animal

of 37

Experiment Committee of Hokkaido University. Pregnant Wister rats were decapitated following ether

anesthesia. Dissociated hippocampal neurons were prepared at embryonic day 18, and were grown on a

homemade 35-mm glass bottom dish containing MEM with 2% B27 Supplement (GIBCO), 2 mM

glutamine, 1 mM sodium pyruvate (GIBCO), penicillin-G potassium salt (50 units/mL), and

streptomycin sulfate (50 μg/mL). On the 4th day in vitro (DIV-4), half of the culture medium was

replaced with MEM containing 2% FBS, 1×N2 Supplement (Invitrogen), and penicillin-streptomycin.

Neuronal cells were transfected at DIV-7 with pcDNA3-GCaMP3, pcDNA3-G-GECO1 or pcDNA3-R-

GECO1 by using calcium phosphate precipitation. Cells were imaged 2-4 days after transfection on a

wide-field epifluorescence inverted microscope (Eclipse Ti-E, Nikon) equipped with a 60× oil

objective lens (Plan Apo NA1.4). For excitation of G-GECO1 and GCaMP3, the samples were

illuminated with light from a 100 W mercury arc lamp that was passed through 25% and 12.5% neutral

density filters and a 480/40 nm bandpass filter. For excitation of R-GECO1, the light was passed

through 25% and 50% neutral density filters and 562/40 bandpass filter. The emission filter for G-

GECO1 and GCaMP3 was 535/40 nm and for R-GECO1 it was 624/40 nm. Digital images were

acquired every 100 ms using an EM-CCD camera (Evolve, Roper) in a 2×2 binning mode.

Preparation and imaging of transgenic C. elegans

To construct the GEM-GECO1 expression plasmid, the GEM-GECO1 coding region (starting from

Met36, as numbered in Fig. S2), was amplified using primers GCaMP_FW_Ce_BamH1 and

GCaMP_RV_Ce EcoR1 and used to construct a destination vector using the Gateway system

(Invitrogen). This was used to make an expression plasmid fragment by LR reaction with an entry

vector containing an odr-10 promoter. Transgenic animals were constructed by microinjection of the

DNA mixture (100 ng/μL podr-10::GEM-GECO1 plasmid with 100 ng/μL pbLH98, a rescuing lin-15

construct) (13).

of 37

Imaging in C. elegans was carried out as previously described (13), with a slight modification. A

worm expressing GEM-GECO1 in the AWA neurons was put into a PDMS microfluidic chip for Ca2+

imaging. For imaging of Ca2+ response of AWA to odorant, diacetyl (0.02%) was diluted with the

imaging buffer (50 mM NaCl, 1 mM CaCl2, 1 mM MgCl2, 0.01% gelatin, and 25 mM HEPES, pH 6.0)

and perfused at the top of the worm head. Optical recordings were performed on an OLYMPUS BX53

upright compound microscope fitted with an ORCA D2 CCD camera (Hamamatsu Photonics K.K.)

and an optical block containing the filters as listed in Table S5. Fluorescence images were acquired by

using HCimage software (Hamamatsu Photonics K.K.) at 1~2 frames/second with a 2×2 binning mode.

Co-imaging of R-GECO1 and a FRET-based ATP indicator

For co-imaging of Ca2+ and ATP dynamics, HeLa cells were transfected with pcDNA3-R-GECO1 and

either pcDNA3-AT1.03 or pcDNA3-mitAT1.03 (mitochondrial targeting). Cell imaging was performed

with an inverted microscope (Eclipse Ti-E, Nikon) equipped with an Orca-R2 digital CCD camera

(Hamamatsu Photonics K.K.), a micro-scanning stage MSS-BT110 (Chuo Precision Industrial Co.

Ltd.), and a Chamlide IC incubator system (Live Cell Instrument). The MetaMorph software package

(Molecular Devices) was used for automated microscope and camera control. Cells were imaged with a

40× oil objective lens (Plan Fluor NA 1.3) and illuminated using a mercury arc lamp thorough 1% and

12.5% ND filters with a 500 ms exposure acquired every 10 s with a 2×2 binning mode for a duration

of 20 min. Imaging was performed at 37 °C using a heated chamber on the microscope stage (INUG2-

ONI, Tokai Hit). Interference filters used for imaging of R-GECO1 and CFP-YFP are provided in

Table S5. Prior to imaging of histamine-induced Ca2+ and ATP dynamics, HeLa cells were washed

twice with DMEM supplemented with 20 mM HEPES, 5.5 mM galactose, and 1 mM pyruvate adjusted

to pH 7.3.

Construction of structural models of GECOs

Models of G-, B-, GEM-, and GEX-GECOs (Figs. S3 and S4) were constructed with initial coordinates

of 37

taken from the atomic structure of Ca2+-bound GCaMP2 (PDB ID 3EVR) (6). In the case of B-GECO,

PyMOL (www.pymol.org) was used to substitute the GCaMP2 chromophore with the blue FP

chromophore (PDB ID 1BFP) through alignment of the α-carbon atoms. The Rosetta fixed backbone

design protocol was used to introduce substitutions and repack amino acid side chains for each GECO

(21). For each substitution, the residue was repacked and minimized on the fixed backbone coordinates

of GCaMP2. The energy was then evaluated using Rosetta all-atom energy function.

The R-GECO1 model was constructed with initial coordinates taken from the atomic structures

of mCherry (PDB ID 2H5Q) and GCaMP2. The model of the hybrid protein was generated in PyMOL

by alignment of the α-carbon atoms of mCherry and the FP portion of GCaMP2. Rosetta molecular

modeling software was then used to link both domains using the kinematic closure loop modeling

protocol (22). Loop residues P60 to V63 (linking M13 and the FP domain) and M300 to T303 (linking

FP and calmodulin) were given full backbone and sidechain-rotamer flexibility to link both domains.

The lowest energy construct was then selected as the initial input for fixed backbone design where the

substitutions unique to R-GECO were made as described for the other GECOs. Output from fixed

backbone design for all GECOs was used for generation of Figs. S3 and S4 using PyMOL.

of 37

SOM text

A colony based screen for Ca2+-dependent fluorescence changes

To facilitate the discovery and optimization of improved GCaMP variants, we developed a system that

enables us to perform an image-based screen for Ca2+-dependent changes of GCaMP fluorescence in

colonies of E. coli. The key to this system is that the GCaMP is targeted to the E. coli periplasm (17),

where the initially high concentration Ca2+ can be experimentally depleted by treatment with a solution

of the high affinity Ca2+-chelator, EGTA. It should be noted that screening libraries of cytoplasmic

GCaMPs expressed in colonies of E. coli has two major drawbacks. The first is that the Ca2+

concentration cannot be readily manipulated. The second is that the intracellular Ca2+ concentration is

approximately 100 nM (23), well below the 542 nM Kd’ of GCaMP3 (9). Accordingly, GCaMP-type

indicators in the E. coli cytoplasm will exist almost entirely in the Ca2+-free state and performing

directed evolution by picking the brightest colonies of E. coli would be counterproductive, as it would

favor variants with increased Ca2+-free state brightness and correspondingly decreased dynamic range.

Picking the dimmest colonies is also counterproductive, since it is much more likely that dim

fluorescence will arise due to poor expression or poor folding (an outcome that could result from many

possible mutations), rather than diminished brightness of the Ca2+-free state relative to the Ca2+-bound

state (the outcome of only a few rare mutations).

In designing our TorA protein export plasmid (pTorPE), we choose to start from the pBAD/His

B (Invitrogen) backbone since the araB promotor provides tightly regulated L-arabinose-dependent

protein expression in E. coli. GCaMP3, the most recent addition to the GCaMP-lineage (9), was used

for validation experiments and as the initial template for directed evolution. The signal peptide of

TMAO reductase (TorA) was chosen as the export sequence since it has previously been shown to

mediate the export of GFP to the periplasm (17). In an effort to minimize the contribution of residual

cytoplasmic protein to the overall fluorescence of each colony, we introduced a C-terminal SsrA

of 37

sequence (AANDENYALAA) that could induce degradation of cytoplasmic proteins by ClpXP and

ClpAP (24). During G-GECO evolution we discovered several variants that exhibited more efficient

export to the periplasm but contained no mutations in the GECO domain. Some of these clones

contained substitutions in the TorA sequence and others had a truncation of the C-terminal SsrA

sequence. For all subsequent rounds of screening, we used a version of the pTorPE plasmid that

contained an enhanced TorA sequence and no SsrA sequence. The sequence of the enhanced TorA

export sequence is MNNNDLFQASRRRFLAQLG[G to S]LT[V to D]AG[M to

T]LGPSLLTPRRATAAQAATDAS. Substitutions relative to the original sequence are in brackets.

Our first proof-of-principle experiment was to spray a fine mist of a pH buffered solution of

EGTA (2 mM EGTA, 30 mM Tris-HCl, pH 7.4) on a plate of E. coli colonies that had been previously

transformed with pTorPE-GCaMP3 and grown on 0.0004% L-arabinose (weight/volume). Rewardingly,

we observed that the fluorescence intensity of the colonies decreased by about 30% upon application of

the EGTA solution. Further investigation revealed that the concentration of free Ca2+ in our LB media

was 156 μM and that application of the EGTA spray was sufficient to diminish free Ca2+ to a

concentration (~300 nM) that was below the GCaMP3 Kd’ and thus shift it primarily to the Ca2+-free

state. Having demonstrated that it was possible to experimentally-induce Ca2+-dependent changes in

GCaMP3 fluorescence in colonies of E. coli, we set out to systematically optimize the concentration of

L-arabinose present in the growth media. A series of agar plates with concentrations of L-arabinose

ranging from 0.0002% to 0.0128% were prepared. E coli was transformed with pTorPE-GCaMP3 and

grown on plates for 20 h at 37 °C, followed by 24 h at room temperature. A digital fluorescence image

(filters information provided in Table S5) of each plate was acquired and then each plate was evenly

and identically sprayed with an EGTA solution using a conventional fine mist sprayer. The EGTA

solution was allowed to soak into the media until no drops were visible and this procedure of spraying

and drying was repeated two additional times. A second digital fluorescence image was acquired and

of 37

then the two images (before and after spraying) were digitally aligned and processed using Image-Pro

Plus (Media Cybernetics) macros. The processing macro calculated the average fluorescence intensities

of every colony in both the first (ICa) and the second image (IEGTA) and exported the values to an

spreadsheet where the fluorescence change (ΔI = (ICa - IEGTA)/ ICa) was calculated for each colony. As

shown in Fig. S8A, ICa reached a maximum around 0.0032% L-arabinose while ΔI decreased

substantially at concentrations of L-arabinose greater than 0.0016%. In addition, we observed

morphological changes in some colonies at concentrations of L-arabinose greater than 0.0032%. Based

on these results we used 0.0002% to 0.0008% L-arabinose for all subsequent experiments.

To ensure that the export system was working as it had had been designed to do, we performed

control experiments in which either buffered solution lacking EGTA was sprayed on the plate or in

which GCaMP3 was confined to the cytoplasm (Fig. S8B). These control experiments revealed that

both export to the periplasm and the application of EGTA were necessary in order to achieve a robust

Ca2+-dependent change in fluorescence intensity. Specifically, when using the export system we

obtained of ΔI values of 0.24 ± 0.044. The value ΔI was reduced to close to the level of the noise if

buffer lacking EGTA was sprayed on identically treated colonies (0.019 ± 0.017) or if GCaMP3 was

expressed without the TorA tag (0.038 ± 0.024).

To augment our primary colony-based screen, we sought to add a secondary screening step in

which the Ca2+-dependent change in fluorescence of each of the top variants was assessed using protein

extracted from the periplasmic fraction of E. coli that had been grown in liquid culture overnight. To

test the effectiveness of this procedure, the dynamic ranges of GCaMP3 in both the periplasmic

fraction (osmotic shock fluid) and the cytoplasmic fraction (B-PER (Pierce) extraction) were tested

using a fluorescence microplate reader. This test revealed that the dynamic range of GCaMP3 was

about 330% in osmotic shock fluid and about 180% in B-PER (Fig. S8C). We attribute the increased

of 37

dynamic range in the osmotic shock fluid to the decreased autofluorescence relative to B-PER. It

should be noted that the osmotic shock fluid contains 5 mM Mg2+ so variants with poor selectivity for

Ca2+ relative to Mg2+ could be identified and discarded at this step. Accordingly, testing of the osmotic

shock fluid was used as our secondary assay during all library screening.

Comparing the properties of GECO variants to small molecule indicators

Despite the widespread availability of genetically encoded protein-based Ca2+ indicators, synthetic

organic dye-based Ca2+ indicators remain popular tools in live cell imaging applications (1, 19). The

ongoing popularity of dye-based Ca2+ indicators is attributable to several factors which include: the

wide selection of fluorescent hues; their high intensiometric and ratiometric signal changes; and a

simple loading procedure that involves incubating tissue with a membrane-permeant form of the

indicator. Disadvantages of dye-based indicators, relative to protein-based indicators, include: the lack

of a straightforward method for targeting to specific subcellular localizations; a more invasive loading

procedure (i.e., loading of membrane permeable dyes vs. transfection); and, obviously, the inability to

create stable cell lines or transgenic animals that permanently express the indicator.

The GECO series of Ca2+ indicators has effectively narrowed the gap, at least in terms of color

selection and the signal changes, between the synthetic dye-based indicators and the genetically

encoded indicators. Indeed, several of the new GECO variants have spectral properties (Table S2) that,

ostensibly, resemble those of popular dye-based indicators. Specifically, parallels can be drawn

between: the G-GECOs and fluo-3; GEX-GECO1 and fura-2; GEM-GECO1 and indo-1; and R-

GECO1 and rhod-2. Fluo-3, like the G-GECOs, is a green intensiometric indicator, with a ~100× signal

change (G-GECOs, ~25×) and a Kd’ of ~400 nM (G-GECOs, ~600-1000 nM). One important

advantage of fluor-3 relative to the G-GECOs is its relatively low pKa of ~5 (G-GECOs, ~7.2-7.6) that

renders it insensitive to physiologically relevant changes in pH. Rhod-2, like R-GECO1, is a red

of 37

intensiometric indicator with a ~200× signal change (R-GECO1, ~16×) and a Kd’ of ~600 nM (R-

GECO1, ~500 nM). Fura-2, like GEX-GECO1, exhibits excitation ratiometric green fluorescence and

binds Ca2+ with a Kd’ of ~100-200 nM (GEX-GECO1, ~300 nM). One disadvantage of fura-2 relative

to GEX-GECO1 is that fura-2 is excited at UV wavelengths (~340 nm and ~380 nm) whereas GEX-

GEXO1 is excited in the visible range (~400 nm and ~480 nm). Of the examples provided here, the

greatest similarity in terms of spectral properties is between indo-1 and GEM-GECO1. Indo-1 is an

emission ratiometric indicator that emits at ~400 nm and ~480 nm (GEM-GECO1, 460 nm and 510 nm)

when excited at ~350 nm (GEM-GECO1, ~400 nm), and that has a Kd’ of ~250 nM (GEM-GECO1,

~340 nM).Values for synthetic dye indicators are from obtained from published literature (1, 19) and

the Molecular Probes Handbook (available online at www.invitrogen.com).

Based on the comparisons provided above, it is apparent that, in terms of the range of hues and

Ca2+ affinities available, genetically encoded indicators may one day surpass synthetic indicators. One

aspect that was not mentioned above were the kinetics of Ca2+ binding and dissociation. The synthetic

indicators discussed above have dissociation rate constants (koff) that are typically 102-104× times faster

(1) than the koff values for the GECOs (Table S3). While modest improvements in the kinetics of

protein-based indicators are certain to be realized, it is unlikely that they could ever rival dye-based

indicators in this respect. However, the slower kinetics of genetically encoded indicators do not hamper

their usefulness in many typical experiments, as we have demonstrated with co-imaging of fura-2 with

either G-GECO1.2 or R-GECO1 (Fig. S9).

Ca2+ dissociation constants and dynamic range of GECO variants

To realize maximum sensitivity, a Ca2+ indicator should have a Kd’ that is between the lowest and

highest physiological Ca2+ concentrations that are expected in a given experiment, and so the Kd’ of a

particular Ca2+ indicator will dictate the type of experiments that it is best suited for (20). The

of 37

concentration of the Ca2+ in the cytoplasm of a resting cell is typically below 100 nM and, following

pharmacological stimulation, can rise to values of 1-10 μM. Accordingly, the GECOs, with their Kd’

values of ~200 nM (B-GECO1), ~300 nM (GEX- and GEM-GECO1), ~500 nM (R-GECO1), ~600

mM (G-GECO1.1), ~700 (G-GECO1), and ~ 1.1 μM (G-GECO1.2), are generally well positioned for

imaging of these changes. More specifically, the indicators with sub μM Kd’ values, such as B-GECO1,

GEX-GECO1, GEM-GECO1, R-GECO1, G-GECO1, and G-GECO1.1, are most sensitive to changes

in the 100 nM to 1 μM range, but are less sensitive to changes in the 1-10 μM range (where G-

GECO1.2 would be preferred), and insensitive to changes above 10 μM. Spontaneous Ca2+ oscillations

(e.g., in neuronal culture) or evoked responses in whole animals (e.g., odor evoked response in C.

elegans) will not generate such high transient Ca2+ concentrations, and so indicators with Kd’ values of

100 nM or less would be ideal. However, for indicators with exceptionally high ratiometric changes,

such as GEM-GECO1, their useful range is broadened to both higher and lower Ca2+ concentrations.

Accordingly, we expect that GEM-GECO1 will be useful for imaging of spontaneous Ca2+ in a wide

variety of contexts and cell types. Future work on optimizing of GECO variants may provide variants

with Kd’ values of 50 nM or less, which should provide improved signal change during spontaneous

and evoked changes in Ca2+ concentration. For imaging of Ca2+ dynamics in the ER, it will be

necessary to engineer GECO variants with Kd’ values in the 10s of μM (20).

of 37

Supporting Figures

Fig. S1. GECO genealogy.

of 37

Fig.

S2.

Seq

uenc

e al

ignm

ent o

f new

var

iant

s des

crib

ed in

this

wor

k. S

ubst

itutio

ns a

re su

mm

ariz

ed in

Tab

le S

1. T

he n

umbe

ring

is

cons

iste

nt w

ith th

at p

revi

ousl

y us

ed fo

r GC

aMP

varia

nts t

hat i

nclu

ded

the

His

-tag-

cont

aini

ng le

ader

sequ

ence

(9).

In th

e cu

rren

t

wor

k th

e le

ader

sequ

ence

was

incl

uded

onl

y fo

r bac

teria

l exp

ress

ion.

For

exp

ress

ion

in c

ell c

ultu

re a

nd C

. ele

gans

, the

lead

er

sequ

ence

was

rem

oved

such

that

M36

was

the

first

resi

due

of th

e pr

otei

n.

of 37

Fig. S3. Location of substitutions in modeled structures of (A) G-GECO1; (B) G-GECO1.1; and (C)

G-GECO1.2. All substitutions are relative to GCaMP3 (9). Three views of each protein are provided,

with substituted residues represented only once within each set of three views. The dotted line

represents the linker that connects the former N- and C-terminus in the cpFP. Note that all three G-

GECOs have substitutions at a residue of CaM that interacts with M13 (E430V) and a residue within

the FP that interacts with the FP to CaM linker (K119I or M). G-GECO1.1 and 1.2 have a third

substitution in close proximity to the M13 to FP linker (N77Y). We speculate that these substitutions

conspire to subtly change the orientation of both M13 and CaM relative to the FP domain, which in

turn modulates the chromophore environment such that its pKa is increased in both the Ca2+-free and

Ca2+-bound states (Table S2). This increase in the pKa diminishes the fluorescence of the chromophore

in the Ca2+-free state by shifting it further toward the neutral (non-fluorescent) form and away from the

anionic (green fluorescent) form.

of 37

Fig. S4. Location of substitutions in modeled structures of (A) B-GECO1; (B) R-GECO1; (C) GEM-

GECO1; and (D) GEX-GECO1. All substitutions are relative to GCaMP3 (9), except those within the

mApple-derived portion of R-GECO1, which are relative to mApple (11). The R377P substitution

present in both B-GECO1 and GEM-GECO1 occurs in an α-helical region of CaM. In the modeled

structure, the proline is in an unrealistic high-energy conformation that clashes with the adjacent

residues in the α-helix. We have chosen to keep the backbone fixed for the models represented here,

although it is apparent that a substantial reorganization of this region of CaM must accompany this

substitution. Detailed structural characterization by X-ray crystallography will be necessary in order to

fully elucidate the effect of R377P and all other substitutions on the structure and mechanism of these

Ca2+ indicators.

of 37

Fig. S5. Additional spectral characterization of GECOs described in this work. For all panels, the Ca2+-

free state is represented with a dotted line and Ca2+-bound state is represented with a solid line. (A)

Absorbance (Abs) and emission spectra (Em) of B-GECO1; (B) Absorbance and emission spectra of

R-GECO1; and (C) Excitation (Ex) and emission spectra of GEX-GECO1.

of 37

Fig. S6. Intensity (or ratio) and dynamic range of GECOs as a function of pH. The dynamic range is

calculated by dividing the intensity (or ratio) of the Ca2+-bound state by the intensity (or ratio) of the

Ca2+-free state. Experimental pKa values are summarized in Table S2. (A) GCaMP3; (B) G-GECO1;

(C) G-GECO1.1; (D) G-GECO1.2; (E) B-GECO1; (F) R-GECO1; (G) GEM-GECO1; and (H) GEX-

GECO1.

of 37

Fig. S7. Stopped flow kinetic characterization of GECOs. Observed relaxation rate constants (kobs) are

plotted as a function of Ca2+ for each of the GECOs described in this work. Association and

dissociation rate constants (kon and koff, respectively) were determined by fitting to the equation kobs

= kon[Ca2+]n + koff. Values are tabulated in Table S3.

of 37

Fig. S8. Validation and optimization of the colony based screen for Ca2+-dependent fluorescence

changes. (A) Brightness and dynamic range as a function of L-arabinose concentration in the growth

media. The smoothed lines are shown to emphasize the trend in each data set. (B) Control experiments

to determine if both export to the periplasm and the application of EGTA are necessary in order to

achieve Ca2+-dependent fluorescent changes in fluorescence intensity. (C) Comparison of the dynamic

range of GCaMP3 extracted from bacteria in osmotic shock fluid and in B-PER.

of 37

Fig. S9. Co-imaging of fura-2 and either R-GECO1 (A) or G-GECO1.2 (B). The slight offset between

peaks for the two indicators arises from the sequential acquisition of images at each time point.

of 37

Supporting Tables Table S1. List of all substitutions for GECOs described in this work

Protein Substitutions relative to GCaMP3 unless otherwise noted

G-GECO0.5 K119I/L173Q/S404G

G-GECO1 K119I/L173Q/S404G/E430V

G-GECO1.1 K69E/N77Y/D86G/K119I/L173Q/K380N/S404G/E430V

G-GECO1.2 K69E/N77Y/D86G/K119I/L173Q/D260G/S404G/E430V

B-GECO0.1 K69E/N77Y/D86G/K119I/L173Q/T223S/Y224H/ R377P/K380Q/

S404G/E430V

B-GECO1 S40P/L60P/D86G/K119I/L173Q/T223S/Y224H/D305G/R377P/K380Q/

S404G

R-GECO1

Substitutions relative to the mApple-derived analogue of GCaMP3:

T47A/L60P/E61V/S63V/E64S/R81G/K83R/Y134C/M158L/N164aD/V228A/

S290P/I366F/K380N/S404G/N414D/E430V

GEM-GECO1 L60P/K69E/N77Y/D86G/N98I/K119I/L173Q/T223S/N302S/R377P/K380Q/S

404G/E430V

GEX-GECO1 K69E/K119I/L173Q/T223S/F315L/F368L/L372Q/K380N/S404G/E430V

flash-pericam1.1 Substitutions relative to flash-pericam:

Q70L/M246L/H304L

flash-pericam1.2 Substitutions relative to flash-pericam:

Q70L/D86G/M246L/H304L

of 37

Table S2. Properties of final GECOs and selected intermediates.

Protein Ca2+

λabs (nm) with ε

(mM-1·cm-1) in

parenthesis

λem with Φ

in

parenthesis

Brightness 1

(mM-1·cm-1) pKa

Intensity

or ratio

change2

± Ca2+

Kd’ for Ca2+

(nM) with Hill

coefficient in

parenthesis

GCaMP3 - 399 (36); 496 (11) 513 (0.20) 2 8.73

12× 542 (2.73) + 399 (20); 496 (50) 513 (0.44) 22 6.6

G-GECO0.5 - 398 (41); 496 (3.4) 512 (0.23) 0.78 10.07

20× 957 (1.98) + 398 (16); 496 (44) 512 (0.40) 17.6 7.19

G-GECO1 - 398 (42); 496 (3) 512 (0.22) 0.66 10.05

25× 749 (2.97) + 398 (25); 496 (41) 512 (0.42) 17.2 7.57

G-GECO1.1 - 398 (42); 496 (3) 512 (0.20) 0.6 10.19

26× 618 (2.05) + 398 (28); 496 (34) 512 (0.46) 15.6 7.5

G-GECO1.2 - 402 (37); 498 (2) 513 (0.25) 0.5 10.44

23× 1150 (2.11) + 402 (22); 498 (33) 513 (0.36) 11.9 7.24

B-GECO0.1 - 380 (14) 446 (0.028) 0.39 5.03

4.3× 260 (2.28) + 380 (14) 446 (0.12) 1.7

B-GECO1 - 378 (22) 446 (0.02) 0.44 5.04

7× 164 (2.64) + 378 (23) 446 (0.18) 4.1 5.60

R-GECO1 - 445 (22); 577 (15) 600 (0.06) 0.72 8.9

16× 482 (2.06) + 445 (9); 561 (51) 589 (0.20) 10.2 6.59

GEM-GECO1 - 397 (34) 511 (0.31) 10.2

6.165 110× 340 (2.94) + 390 (36) 455 (0.18) 6.5

GEX-GECO1

- 397 (32), 482 (0) 512 (0.21) 6.7

~65 26× 318 (2.78) + 390 (32), 482 (2.5)

506 (0.19,

0.39) 3 6.1, 14

1Brightness is defined as the product of ε and Φ. 2Defined as the ratio of emission intensities for intensiometric GECOs, the ratio of emission ratios for GEM-GECO1, and the ratio of excitation ratios for GEX-GECO1, where the Ca2+-bound state is the numerator and the Ca2+-free state is the denominator. For R-GECO1, which undergoes an 11 nm blue shift when it binds Ca2+, the intensity at 589 nm is used for both states. The emission ratio for GEM-GECO1 is defined as (fluorescence intensity for excitation at 390 nm and emission at 455 nm)/(fluorescence intensity for excitation at 390 nm and emission at 511 nm). The excitation ratio for GEX-GECO1 is defined as (fluorescence intensity for excitation at 395 nm and emission at 512 nm)/(fluorescence intensity for excitation at 450 nm and emission at 512 nm). 3Quantum yield of GEX-GECO1 in the Ca2+-bound state was measured for excitation at both 390 nm and 482 nm. 4The brightness of GEX-GECO1 in the Ca2+-bound state for excitation at both 390 nm and 482 nm is provided. 5The pKa of GEM-GECO1 and GEX-GECO1 is the pH at which the dynamic range is 50% of maximum.

of 37

Table S3. Kinetic characterization of final GECOs.

Protein kon (M-ns-1) koff (s-1) n Kd’, kinetic (nM) 1 Kd’, static (nM) 2

GCaMP3 6.26 × 1016 0.700 2.7 618 542

G-GECO1 9.08 × 1015 0.700 2.7 831 749

G-GECO1.1 8.17 × 1015 0.675 2.6 809 618

G-GECO1.2 8.55 × 1017 0.700 3.0 936 1150

B-GECO1 4.68 × 1012 0.490 2.0 324 164

R-GECO1 9.52 × 109 0.752 1.6 484 482

GEX-GECO1 8.14 × 1012 1.030 2.0 356 318

GEM-GECO1 2.58 × 1012 0.224 2.0 295 340

1Kd’, kinetic = (koff/kon)1/n. 2From Table S2

of 37

Table S4. Systematic characterization of the Ca2+-dependent fluorescence of GECOs in

HeLa cells. Cells were treated first with histamine (abb. His), then with Ca2+/ionomycin (abb.

Ca2+), and finally with EGTA/ionomycin (abb. EGTA).

Protein n1 Maximum Ca2+ to

minimum EGTA ratio Maximum His to

minimum His ratio Maximum His to

maximum Ca ratio

GCaMP3 38 5.3±2.3 4.6±0.9 0.71±0.14

G-GECO1 31 9.4±2.6 8.2±4.0 0.64±0.22

G-GECO1.1 38 11.6±2.9 11.7±3.4 0.69±0.22

G-GECO1.2 42 19.0±7.4 10.2±4.7 0.72±0.43

B-GECO1 38 4.2±1.6 3.5±0.9 0.83±0.26

R-GECO1 22 4.9±1.9 9.2±1.3 0.98±0.15

GEM-GECO1 29 28.5±11.7 28.2±14.7 0.86±0.19

GEX-GECO1 50 10.6±4.1 9.7±3.5 1.04±0.40

1Number of individual transfected cells on which systematic calibration experiments were performed.

of 37

Table S5. Filters1 for screening and live cell imaging.

Protein Colony-based

screening HeLa cell culture

Rat hippocampal

neuron culture C. elegans

GCaMP3,

flash-pericam,

G-GECO series

Ex: 470/40 nm

Dichroic: none

Em: 525/50 nm

Ex: 470/40 nm

Dichroic: 505 nm

Em: 510 nm long pass

or 520/35 nm

Ex: 480/40 nm

Dichroic: 505 nm

Em: 535/40 nm

N/A2

B-GECO1

Ex: 395/40 nm

Dichroic: none

Em: 460/40 nm

Ex: 370/36 nm

Dichroic: 409 nm

Em: 400 nm long pass

or 447/60 nm

N/A N/A

R-GECO1

Ex: 535/50 nm

Dichroic: none

Em: 610/75 nm

Ex: 562/40 nm

Dichroic: 593 nm

Em: 624/40 nm

Ex: 562/40 nm

Dichroic: 593 nm

Em: 624/40 nm

N/A

GEM-GECO1

Ex: 395/40 nm

Dichroic: none

Em1: 460/40 nm

Em2: 525/50 nm

Ex: 377/50 nm

Dichroic: 409 nm

Em1: 447/40 nm

Em2: 520/35 nm

N/A

Ex: 414/46 nm

Dichroic: 450 nm

Em1: 483/32 nm

Em2: 542/27 nm

GEX-GECO1

Ex1: 395/40 nm

Ex2: 470/40 nm

Dichroic: none

Em: 535/30 nm

Ex1: 365/50 nm

Ex2: 475/40 nm

Dichroic: 495 nm

Em: 520/35 nm

N/A N/A

CFP-YFP

FRET pair N/A

Ex: 438/24 nm

Dichroic: 458 nm

Em1: 483/32 nm

Em2: 534/30 nm

N/A N/A

Fura-2 N/A

Ex1: 340/26 nm

Ex2: 387/11 nm

Dichroic: 409 nm

Em: 510/84 nm

N/A N/A

1Bandpass filters are defined using the convention of center wavelength/full bandwidth in nm unless otherwise noted.

Filters were purchased from Semrock and Chroma. 2Not applicable.

of 37

Table S6. Oligonucleotides used in this work

Name Sequence (5’ to 3’)

FW_TorA GCGATGCCATGGGTTTAAAGAGGAGAAAGGTCATGAACAATAACGATCTCTTTCAG

RV_TorA ATGATGAGAACCTCTAGAAGCGTCAGTCGCCGCTTG

FW_6His GACGCTTCTAGAGGTTCTCATCATCATCATCATCATGG

RV_SsrA-GCaMP3 GCGATGAAGCTTCTAAGCTGCTAAAGCGTAGTTTTCGTCGTTTGCTGCCCCGGGACCA

CCCTTCGCTGTCATCATTTGTACAAACTCTTCGTAGTTT FW_XbaI-6His GCGATGTCTAGAGGTTCTCATCATCATCATCATCATGGTATGGCTAGC

RV_GCaMP-XmaI GCGATGCCCGGGACCACCCTTCGCTGTCATCATTTGTACAAACTCTTCGTAGTTT

RV_GCaMP-Stop-HindIII GCGATGAAGCTTCTACTTCGCTGTCATCATTTGTACAAACTCTTCGTAGTTT

FW_GCaMP3_c105a GATAAGGATCTCGCCACAATGGTCGACTCATCACG

G-B_V63VILM AGGTCGGCTGAGCTCACTAGAGAACVTVTATATCAAGGCCGACAAG

G-B_T223ST_Y224H CTCGTGACCACCCTGWCNCACGGCGTGCAGTGCT

G-B_R377X_N380X GACTTCCCTGAGTTCCTGACAATGATGGCANNKAAAATGNNKGACACAGACAGTGAAG

AAGAAATTAGAGAA

FW_XhoI-X-148mc GCGATGCTCGAGKBYGAGCGGATGTACCCCGAGGAC

FW_XhoI-X-147mc GCGATCCTCGAGKBYTCCGAGCGGATGTACCCCG

RV_GGTGGS-mCherry ACTCCCGCCTGTACCTCCCTTGTACAGCTCGTCCATGCCGC

RV_MluI-X 144mc GGCATGACGCGTRVMCTCCCAGCCCATGGTCTTCTTCTG

RV_MluI-X 143mc GGCATGACGCGTRVMCCAGCCCATGGTCTTCTTCTGCAT

FW_GGTGGS-mCherry GGGAGGTACAGGCGGGAGTATGGTGAGCAAGGGCGAGGAGGATAA

Destroy_MluI CAAAGCCATGACAAAAACGCGAAACAAAAGTGTCTATAATCAC

GCaMP_FW_BamH1 GAGGATCCACCATGGTCGACTCATCACGTC

GCaMP_RV_EcoR1 CGCGAATTCTTACTTCGCTGTCATCATTTGTAC

GCaMP_RV_HindIII CGCAAGCTTCTACTTCGCTGTCATCATTTGTAC

Cpd_S5P_FW_BamH1 GAGGATCCACCATGGTCGACTCACCACGTC

T7_FW TAATACGACTCACTATAGG

F-EGFP-C CATGGTCCTGCTGGAGTTCGTG

BGH_RV TAGAAGGCACAGTCGAGG

GCaMP_2NLS_FW1 GAAGGTCGATCCTAAGAAGAAACGCAAGGTGATGGTCGACTCATCACGTC

Cpd_2NLS_FW1 GAAGGTCGATCCTAAGAAGAAACGCAAGGTGATGGTCGACTCACCACGTC

Cpd_2NLS_FW2 GAGGATCCACCATGGACCCAAAAAAGAAGCGGAAGGTCGATCCTAAGAAG

GCaMP_FW_BamH1_mito GAGGATCCAACCATGGTCGACTCATCACGTC

Cpd_S5P_FW_BamH1_mito GAGGATCCAACCATGGTCGACTCACCACGTC

GCaMP_FW_Ce_BamH1 CGACGGATCCAAAAAATGGTCGACTCATCACGTCG

GCaMP_RV_Ce EcoR1 CGACGAATTCTTACTTCGCTGTCATCATTTGTAC

of 37

References and Notes 1. R. Y. Tsien, Monitoring cell calcium, in Calcium as a Cellular Regulator, E. Carafoli, C. B. Klee,

Eds. (Oxford University Press, Oxford, 1999), pp. 28–54.

2. A. Miyawaki et al., Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin. Nature 388, 882 (1997). doi:10.1038/42264 Medline

3. J. Nakai, M. Ohkura, K. Imoto, A high signal-to-noise Ca(2+) probe composed of a single green fluorescent protein. Nat. Biotechnol. 19, 137 (2001). doi:10.1038/84397 Medline

4. T. Nagai, A. Sawano, E. S. Park, A. Miyawaki, Circularly permuted green fluorescent proteins engineered to sense Ca2+. Proc. Natl. Acad. Sci. U.S.A. 98, 3197 (2001). doi:10.1073/pnas.051636098 Medline

5. Single-letter abbreviations for amino acid residues are A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.

6. Q. Wang, B. Shui, M. I. Kotlikoff, H. Sondermann, Structural basis for calcium sensing by GCaMP2. Structure 16, 1817 (2008). doi:10.1016/j.str.2008.10.008 Medline

7. J. Akerboom et al., Crystal structures of the GCaMP calcium sensor reveal the mechanism of fluorescence signal change and aid rational design. J. Biol. Chem. 284, 6455 (2009). doi:10.1074/jbc.M807657200 Medline

8. Y. N. Tallini et al., Imaging cellular signals in the heart in vivo: Cardiac expression of the high-signal Ca2+ indicator GCaMP2. Proc. Natl. Acad. Sci. U.S.A. 103, 4753 (2006). doi:10.1073/pnas.0509378103 Medline

9. L. Tian et al., Imaging neural activity in worms, flies and mice with improved GCaMP calcium indicators. Nat. Methods 6, 875 (2009). doi:10.1038/nmeth.1398 Medline

10. Materials and methods are available as supporting material at Science Online..

11. N. C. Shaner et al., Improving the photostability of bright monomeric orange and red fluorescent proteins. Nat. Methods 5, 545 (2008). doi:10.1038/nmeth.1209 Medline

12. H. Zhao, W. Zha, In vitro ‘sexual’ evolution through the PCR-based staggered extension process (StEP). Nat. Protoc. 1, 1865 (2006). doi:10.1038/nprot.2006.309 Medline

13. Y. Shinkai et al., Behavioral choice between conflicting alternatives is regulated by a receptor guanylyl cyclase, GCY-28, and a receptor tyrosine kinase, SCD-2, in AIA interneurons of Caenorhabditis elegans. J. Neurosci. 31, 3007 (2011). doi:10.1523/JNEUROSCI.4691-10.2011 Medline

14. H. Imamura et al., Visualization of ATP levels inside single living cells with fluorescence resonance energy transfer-based genetically encoded indicators. Proc. Natl. Acad. Sci. U.S.A. 106, 15651 (2009). doi:10.1073/pnas.0904764106 Medline

15. M. Nakano, H. Imamura, T. Nagai, H. Noji, Ca²⁺ regulation of mitochondrial ATP synthesis visualized at the single cell level. ACS Chem. Biol. 6, 709 (2011). doi:10.1021/cb100313n Medline

16. F. Zhang et al., Optogenetic interrogation of neural circuits: Technology for probing mammalian brain structures. Nat. Protoc. 5, 439 (2010). doi:10.1038/nprot.2009.226 Medline

of 37

17. C. M. Barrett, N. Ray, J. D. Thomas, C. Robinson, A. Bolhuis, Quantitative export of a reporter protein, GFP, by the twin-arginine translocation pathway in Escherichia coli. Biochem. Biophys. Res. Commun. 304, 279 (2003). doi:10.1016/S0006-291X(03)00583-7 Medline

18. Z. Cheng, R. E. Campbell, Assessing the structural stability of designed beta-hairpin peptides in the cytoplasm of live cells. ChemBioChem 7, 1147 (2006). doi:10.1002/cbic.200500540 Medline

19. G. Grynkiewicz, M. Poenie, R. Y. Tsien, A new generation of Ca2+ indicators with greatly improved fluorescence properties. J. Biol. Chem. 260, 3440 (1985). Medline

20. A. E. Palmer, R. Y. Tsien, Measuring calcium signaling using genetically targetable fluorescent indicators. Nat. Protoc. 1, 1057 (2006). doi:10.1038/nprot.2006.172 Medline

21. R. Das, D. Baker, Macromolecular modeling with Rosetta. Annu. Rev. Biochem. 77, 363 (2008). doi:10.1146/annurev.biochem.77.062906.171838 Medline

22. D. J. Mandell, E. A. Coutsias, T. Kortemme, Sub-angstrom accuracy in protein loop reconstruction by robotics-inspired conformational sampling. Nat. Methods 6, 551 (2009). doi:10.1038/nmeth0809-551 Medline

23. P. Gangola, B. P. Rosen, Maintenance of intracellular calcium in Escherichia coli. J. Biol. Chem. 262, 12570 (1987). Medline

24. A. W. Karzai, E. D. Roche, R. T. Sauer, The SsrA-SmpB system for protein tagging, directed degradation and ribosome rescue. Nat. Struct. Biol. 7, 449 (2000). doi:10.1038/75843 Medline

supporting online material for -...

Documents